Fenvi FU-AX1800 photos and information dumps #309
Comments
|
My adapter has a MAC address beginning with |
|
This adapter feels extremely cheaply made, to the point that I'm not even certain that the "antennas" aren't placebos, but on the other hand it seems to have ridiculously good receive reception, to the point that it's successfully decoding beacons from neighbors' networks that I've never seen with any other adapter. I've summarized the scan results of my own networks in the table below, comparing receive signal strengths with an old AR9271-based 802.11n USB stick against those with the Fenvi FU-AX1800.
|
|
I bought 3 (for tri-band) for a total of 25usd (aliexpress promo going on) to re-start my rpi portable travel router project, hopefully this will work (unlike those pesky ANDDEAR 7612U adapters that didn't work last time) hopefully I can run 6ghz AP 80mhz too
In my head the rpi4 could do: built in 2.4ghz, usb1 5ghz, usb2 6ghz, usb3 backhaul (on whatever the existing wifi frequency uses) |
My comment here is separate but somewhat related. I have done a few tests with my mt7921au adapters and other adapters. Maybe it is the part of the universe I live in or something I don't understand but even when I see the mt7921au adapters with the same or slightly lower signal reading, they do as good or better when throughput is tested in the same situation. Thoughts?
Also on this topic. A few years ago I bought an Alfa ACHM. I used it for a few months and then one day I was doing a range test with several adapters. I had to redo things several times because I could believe how the ACHM showed range that was so much longer than the other adapters. The Alfa ACH was second place. Alfa does list these two adapter as "MAX Power" so I guess they are putting good quality anteenas and amps in them. Keep that info coming. We are interested. Thanks. |
That is hard to believe. Where is my usb power meter? I have 2 adapters with the mt7921au chipset and I know how to torture an adapter and I just can't heat them up. How did Mediatek do this? |
I don't know if that mediatek for their phones has tsmc fabbed 4nm SoCs for their dimensity line which are also very power efficient and have high performance (comparable to flagship qualcomm snapdragon SoCs) so it would not surprise me if usb wifi adapters are made with the same/similar fab process for the transistors that are being used SoCs for phones include: CPU, RAM, wifi card maybe not leading edge, but 1 or 2 generations behind (since these adapters are so cheap after all, but I think the amount of silicon these usb sticks need is also much lower than a cpu, for example) If we are getting better process nodes for denser/more efficient transistors by using more up to date lithography processes then of course we will get much better efficiency and lower heat. Have you ever used a AMD 7nm laptop compared to an intel 14nm laptop? The difference is huge in efficiency and performance. wikidevi or techinfodepot don't always list the lithography process used but here is an example: https://www.pcguide.com/news/mediatek-filogic-380-chip-promises-next-gen-wi-fi-7-connectivity/
According to wikidevi:
So if Filogic 380 which is wifi7 is using a 6nm lithography process that leads to the question, what fab process is the filogic 320 using? likely much newer and several generations newer than whatever usb stick you have been using so far :) Edit: I suspect filogic 330/320 uses 12nm lithography https://corp.mediatek.com/news-events/press-releases/mediatek-announces-filogic-connectivity-family-with-new-filogic-830-and-filogic-630-wi-fi-6-6e-chips because filogic 830/630 also use 12nm lithography Edit2: @morrownr I did some more digging: according to https://wikidevi.wi-cat.ru/MediaTek#ac_.28Wi-Fi_5.29 the MT7610U chipset inside your AWUS036ACHM adapter is of the MT7615 generation/family of devices. If I look up https://wikidevi.wi-cat.ru/MediaTek_MT7615, it says
TSMC 12nm is roughly equivalent to intel 14nm transistor density now you know why: so going from MT7610U -> MT7921AU is a 3 generation jump in efficiency/processing power with each generation jump being a ~20-30% uplift, so 1.25^3 = 1.95x more efficient/processing power roughly |
|
Hi @Snuupy Thanks for the info. I finally found my usb power meter! Here are the readings: Adapter: Alfa AXML When surfing, the numbers sometimes got up to around 100-120 mA but surfing is not really so stressful. Those are good numbers for those needing to limit power usege and will certainly be welcomed by the RasPi users but it seems to indicate the 100 mA figure provided by lsusb is not entirely accurate. The thermal readings I have seen with my 2 adapters that use the mt7921au chipset are very impressive. I test with extended periods using iperf3 and I have seen no thermal problems at all. This is the same test that will allow adapters with some chipsets and limited ventilation to cook an egg. |
|
for the power meter does data transfer as normal? or do you have to get a specific model? looks like with 1.2A we can get 3 or 4 simultaneous adapters :D |
|
I have two of these and both work on my Raspberry Pi 4. Here is one taken apart: List of all AC caps from my version of iw: |
|
I guess we know the foldable antennas aren't placebos now! 📶 Thanks for the info, @Anthony96922! |
|
Hi. I'm looking into buying this for my (ax) penetration testing needs on irregular basis, learning mainly so 2dbi antenna will be fine, but I'm wondering about modes needed for that, driver issues or hardware faults at this point on this unit. Have you tested monitor mode, injection and stability whit airodump-ng? |
|
It appears to support monitor mode based on the output iw list. |
|
I have two adapters with the m7921au chipset. I have done basic monitor mode, injection and deauth testing and did not see any problems. I remember some reports from around July to September of 2022, not long after the adapters with this chipset start to ship. I remember one guy that does a lot of monitor mode work that said the driver works well in monitor mode. He said that it worked well in 6 Ghz. A lot of Linux users now have adapters with this chipset and I am not seeing any complaints about monitor mode. |
|
Tnx both, I already ordered this unit as it was in "super deal" and "3 units left" state, for 13eur not that much to loose. PS: Great repo, lots of useful information about this topic. If this chipset is very good will order some hi gain, better quality unit. |
Thanks. Something like this site needed to happen and I get a lot of help from users. If you have time to do a review once you have used this adapter, people would appreciate it. I'll put it with the posting in the Plug and Play List.
The Alfa adapters in the Plug and Play list for this chipset both have removal antennas which opens up some options. We don't have a Alfa "MAX Power" adapter yet but hopefully one comes. We do have at least one, maybe two additional chipsets coming for USB adapters at some point in the next year. The new Mediatek mt7925 chipset driver is starting to go into the kernel with kernel 6.7. It is a full up WiFi 7. Another driver in the kernel before product is available. Woo Hoo! We could get the mt7922au chipset before that. It is WiFi 6e like the mt7921au but has 160 Mhz channel width capability. USB Wifi just keeps getting better on Linux. |
As I will be using it for penetration testing and stuff like that I would not be able to comment on regular daily usage. But I can do speedtests and overall impression review, I have ax APs at home. I would not rush to buy wifi7 at this point as standard is not finalized, but good to see timely development from vendors. |
Anything you can add is appreciated.
I don't think we will have the opportunity to rush out and buy new WiFi 7 USB adapters very soon as USB adapters always come after other products such as cards but we are likely looking at a situation like with the mt7921au chip in that the the driver will be in the kernel for a few months before the prodducts hit the shelves. I think the first time that ever happened was with the mt7921au. I've used usb wifi adapters for many years on Linux and it is the first time I remember it happening. Thanks Mediatek for making it happen. |
Solved here #319 (comment)I asked in another issue but just in case I will mentioned here as well. What is the maximum speed you guys managed to reach with the device? I got mine delivered today and I can't get it above 30 mbit/s. It almost feels like it's capped somehow! I am connecting to 5Ghz channel I am sure because I have different WIFI names for 2.4Ghz and 5Ghz Just in case it useful here is my system information: |
|
Hi, does anyone have an FCC ID or schematic? |
|
just a note that the FU-1800 (and most other adapters of this type) have a maximum transmission signal strength of 20 dBm I saw in the spec sheet so they will never be as good as an actual AP that can follow regdb guidelines of 27 dBm for example for $8? it's fine but any more than that I'd start considering a real AP (or something with pcie) instead. |
Manufacturer's product page: https://fenvi.com/product_detail_44.html
Purchased from: https://www.aliexpress.com/item/1005005935638503.html
After loading the
USregulatory domain:Relevant excerpts from kernel log (Linux 6.1.52+):
The text was updated successfully, but these errors were encountered: